Why We're Polarized

by Ezra Klein

I think we are, or we can be. But toxic systems compromise good individuals with ease. They do so not by demanding we betray our values but by enlisting our values such that we betray each other. What is rational and even moral for us to do individually becomes destructive when done collectively.

Systems thinking, he writes, “is about understanding how accidents can happen when no parts are broken, or no parts are seen as broken.”

The American political system—which includes everyone from voters to journalists to the president—is full of rational actors making rational decisions given the incentives they face. We are a collection of functional parts whose efforts combine into a dysfunctional whole.

With a quick sleight of hand, identity becomes something that only marginalized groups have.

The most powerful identities in modern politics are our political identities, which have come, in recent decades, to encompass and amplify a range of other central identities as well. Over the past fifty years, our partisan identities have merged with our racial, religious, geographic, ideological, and cultural identities. Those merged identities have attained a weight that is breaking our institutions and tearing at the bonds that hold this country together. This is the form of identity politics most prevalent in our country, and most in need of interrogation.

Party affiliation was a tautology for itself, not a rich signifier of principles and perspective.

Kennedy warned that “the country was already split vertically, between sections, races, and ethnic groups,” so it would be “dangerous to split it horizontally, too, between liberals and conservatives.”6 Politics, in this telling, was meant to calm our divisions, not represent them.

when Gerald Ford ran against Jimmy Carter, only 54 percent of the electorate believed the Republican Party was more conservative than the Democratic Party. Almost 30 percent said there was no ideological difference at all between the two parties.10 Imagine, in a world where the ideological difference between the Democratic and Republican Parties was slim enough to confuse half the population, how much less force party identity must have carried.

today’s independents vote more predictably for one party over the other than yesteryear’s partisans.

self-described independents who tended to vote for one party or the other were driven more by negative motivations.

What has happened to American politics in recent decades is that the parties have become visibly, undeniably more different, and the country has rationally become more partisan in response.

“Across 10 measures that Pew Research Center has tracked on the same surveys since 1994, the average partisan gap has increased from 15 percentage points to 36 points.”18 It’s worth being clear about what this means: if you’re a Democrat, the Republican Party of 2017 poses a much sharper threat to your vision of a good society than the Republican Party of 1994 did.

To put that more simply, a voter who mostly ignores American politics today is clearer on the differences between the two parties than political junkies and partisan loyalists were in 1980.

In 2014, Pew found that 37 percent of Republicans and 31 percent of Democrats viewed the other party as “a threat to the nation’s well-being.” By 2016, that was up to 45 percent of Republicans and 41 percent of Democrats.24

In 1953, in the so-called Black Belt—the region of Alabama where the black population exceeded the white population—“only 1.3 percent of eligible African Americans were registered. Two counties had no black voters whatsoever.”8

Abraham Lincoln was the first Republican president; the South’s enmity toward the Republican Party was thus signed in blood. The Democratic Party supported redistribution from the rich to the poor—and the North was rich and the South was poor.

1964 Civil Rights Act. As Geoffrey Kabaservice shows in Rule and Ruin, his history of Republican moderation, “eighty percent of House Republicans supported the bill, as opposed to sixty percent of House Democrats.”

In the end, “as with the House vote, a greater proportion of Senate Republicans than Democrats voted for cloture and passage of the [Civil Rights Act]: more than four-fifths of the Republicans but only some two-thirds of the Democrats,” wrote Kabaservice. So why are the Democrats seen as the party that passed the Civil Rights Act? There, the answer is simple. Because they were the party that passed the Civil Rights Act. They held the majority in both chambers and the presidency. They chose to snap their alliance with the Dixiecrats to pursue justice. Bill Moyers, who served as special assistant to Johnson, recalls finding the president brooding in his bedroom the night he signed the Civil Rights Act. “I think we just delivered the South to the Republican Party for a long time to come,” Moyers remembers Johnson saying.

That, then, is the story of the long period of depolarization in American politics. The South was in the Democratic Party, but it didn’t agree with the Democratic Party—particularly once liberalism’s vision of redistribution and uplift expanded to include African Americans. So southern Democrats had ideological reasons to compromise with Republicans but political reasons to compromise with national Democrats. Southern power kept the Democratic Party less liberal than it otherwise would’ve been, the Republican Party congressionally weaker than it otherwise would’ve been, and stopped the two parties from sorting themselves around the deepest political cleavage of the age.

is not that American politics was not riven by sharp, even violent disagreement in this era; it’s simply that these fights did not map cleanly onto party.

The country holds the same mix of beliefs about pot in both examples. It’s just that in the second example, those beliefs are sorted by party.

That’s polarization: the opinions themselves changed to cluster around two poles, with no one left in the middle.

sorting is just a subcategory of polarization.16 In practical terms, he writes, both of these “have the consequence of increasing the tension between the two ends of the spectrum,” which is what polarization is meant to describe.

The polarization versus sorting debate is better understood as describing issue-based polarization an...

Crucially, these forms of polarization reinforce each other. Issue-based polarization leads to political identity polarization: if

as intense polarization around the issue of civil rights drove party polarization around civil rights.

Polarization begets polarization. But it doesn’t beget extremism. We often assume that voters and political systems that split the difference are less extreme than those that don’t, but this idea proves incoherent upon a moment’s inspection.

Obamacare was a public-private system with Republican roots that paid for itself through a mixture of tax increases and spending cuts, while Medicare was a liberal government takeover of health care for the elderly that created an open-ended entitlement with no dedicated way to pay its full costs. And that’s assuming ideological extremism is an idea with internal logic in the first place, which I also doubt.

in the era when Washington was least polarized, political consensus rested on a foundation of racial bigotry that most would find abhorrent today. The compromises Congress made to preserve the peace included voting down anti-lynching laws and agreeing to lock most African Americans out of Social Security. I would call that political system far more ideologically extreme than the one we have today, even as it was less polarized. Political scientists agree that the mid-twentieth century was the low ebb of political polarization, particularly in Congress. But the mid-twentieth century was not an era in which the world outside Washington was either serene or moderate. This

they’re actually quite extreme.”17 When polarization is driven by allegiance to political parties, it can be moderating. Political parties want to win elections, so they try to champion ideas that won’t get their candidates crushed at the ballot box. People who aren’t attached to one party or the other are free to hold much more unpopular opinions.

When I say the political coalitions are becoming more sorted and more polarized, I mean only that: there is less ideological overlap, fewer of us are caught in the middle, and there is more tension between the poles. Nothing about those dynamics makes the opinions partisans hold in 2020 more extreme than those held by their forebears. Banal views held widely in that era would get you run out of polite society today, and rightly so.

The death of the Dixiecrats cleared the way for southern conservatives to join the Republican Party and northern liberals to join the Democratic Party. That let the parties sort themselves ideologically, such that there are no longer any House Democrats more conservative than any House Republicans

the 1952 presidential election, the American National Election Survey found that 6 percent of self-identified Democrats and 2 percent of self-identified Republicans were nonwhite. In 2012, the same survey found 43 percent of self-identified Democrats, but only 9 percent of self-identified Republicans, were nonwhite.19 So not only was the 2012 electorate far, far more racially diverse than the 1952 electorate, but that diversity was concentrated in the Democratic Party.

In 2014, Pew reported that the single largest religious group in the Republican coalition was evangelical Protestants. And the Democrats? Their single largest religious group was the religiously unaffiliated, the “nones.”

How Democracies Die, “the two parties are now divided over race and religion—two deeply polarizing issues that tend to generate greater intolerance and hostility than traditional policy issues such as taxes and government spending.” I’d amend that slightly: the parties are dividing over fundamental identities that tend to generate intolerance and hostility, and the issue conflicts are just one expression of that division.

From 1972 to 1984, the average difference between how a state voted in one presidential election and how it voted in the next was 7.7 percentage points. From 2000 to 2012, it was only 1.9 percentage points. We are fixed in political place.

In less than twenty-five years, the percentage of voters who lived in a district where almost everyone thought like them politically went from 1 in 20 to 1 in 5.

Tucked inside these numbers is a growing urban-rural divide. There is no dense city in America that routinely votes Republican. There are few rural areas that vote Democratic.

nine hundred people per square mile: above that, areas trend Democratic; below it, they turn Republican.23

in both 1992 and 1996, Bill Clinton “carried nearly half of America’s 3,100 counties. But since then, Democrats have retreated into the nation’s urban centers.”24 In 2000, Al Gore won the popular vote with fewer than seven hundred counties. In 2012, Obama won the popular vote by much more than Gore, but carried only about six hundred counties. In 2016, Hillary Clinton won the popular vote with fewer than five hundred counties—more

House Democrats now represented 78 percent of all Whole Foods locations, but only 27 percent of Cracker Barrels.

dominant considerations when people are choosing a place to move are housing prices, school quality, crime rates, and similar quality-of-life questions.

In 2017, Pew found that “most Republicans (65%) say they would rather live in a community where houses are larger and farther apart and where schools and shopping are not nearby. A majority of Democrats (61%) prefer smaller houses within walking distance of schools and shopping.”30 Thus, a preference that seems nonpolitical on its face—“I want a big house with a yard,” or “I want to live in a diverse city with lots of new restaurants”—becomes yet another force pulling partisans away from each other.

Psychologists speak of the Big Five personality traits: openness to experience, conscientiousness, extroversion-introversion, agreeableness, and neuroticism.

“Democrats and Republicans are now sharply distinguished by a set of basic psychological dispositions related to experiential openness—a general dimension of personality tapping tolerance for threat and uncertainty in one’s environment.”31

one is more fundamental than any other in determining which side of the divide you gravitate toward: your perception of how dangerous the world is.

openness to experience—and the basic optimism that drives it—is associated with liberalism, while conscientiousness, a preference for order and tradition that breeds a skepticism toward disruptive change, connects to conservatism.

Those open to new experiences are not just hanging Jackson Pollock prints in disorganized bedrooms while listening to techno-pop reinterpretations of Bach by experimental jazz bands. They are also more likely to identify themselves as liberals.

Their choices map onto our politics not because they are trying to serve one side of the political divide but because our politics map onto our deeper preferences, and those deeper preferences drive much more than just our politics. We like to think that we choose our politics by slowly, methodically developing a worldview, using that worldview to generate conclusions about ideal tax and health and foreign policy, and then selecting the political party that fits best. That’s not how the political psychologists see it. They argue that our politics, much like our interest in travel and spicy food and being in crowds, emerges from our psychological makeup. “Certain ideas are attractive to some people and repulsive to others, and that means, essentially, that ideologies and psychologies are magnetically drawn to each other,”34

the job of the conservative, wrote National Review founder William F. Buckley, is to “[stand] athwart history, yelling Stop.”